MEASURING THE PERFORMANCE OF ASSETS: A REVIEW OF THE FACILITY CONDITION INDEX

Measurement of performances in asset management is an enabling factor to define an effective strategy. The scope of this research concerns the investigation on the Facility Condition Index (FCI): one of the most common metrics related to the built environment. The research aims at providing a critical review concerning the FCI and its use. The research methodology has been developed according to three steps: bibliographic research, bibliometric analysis and critical review. Most of the sources agree that FCI provides a reliable economic measurement of maintenance needs, though it must be tied to a strong condition assessment methodology. Moreover, different methodologies for computation of the deferred maintenance and the current replacement value can be found in literature and many researchers emphasised the limits of FCI when dealing with components characterised by a considerable difference in replacement costs. In conclusion, FCI is included in the wider framework of asset management business process.


Introduction
Performance measurement in asset management, since the last few decades, has gained momentum as a powerful means for decision making (Alexander, 1992). Performance measurement can be intended both as an enabling factor for the successful achievement of the objectives of an organisation and as a learning process within the company  for reaching a higher level of effectiveness in decision making and delivery of better services (Yang, Yeung, A. P. C. Chan, Chiang, & D. W. M. Chan, 2010). The use of Key Performance Indicators (KPIs) eases decision making, allowing to synthesise phenomena related to the built environment, without losing the systemic value of information (Ladiana, 2007). According to the level of aggregation, information acquires different meanings corresponding to the management needs. In Facility Management (FM), for instance, three decision making levels corresponding to specific sets of KPIs can be identified: strategic, tactical and operational (Ente Italiano di Unificazione, 2018; . Performance measurement is always tied to an assessment process, which can exploit a wide combination of metrics belonging to different disciplinary fields (Ama-ratunga, Baldry, & Sarshar, 2000;Amaratunga & Baldry, 2003). These metrics can be collected in a performance measurement framework which varies according to the phenomena to be measured (Douglas, 1996). In Architecture, Engineering, Construction and Operation (AECO), KPIs are exploited for measurement of performances mainly related to technical, functional and economic/financial issues (Amaratunga & Baldry, 2003). Even if it has been demonstrated that economic and financial indicators are too narrow to achieve a comprehensive knowledge of assets' performances (Pärn, Edwards, & Sing, 2017;Yang et al., 2010), they are widely used in Asset Management (AM). Thus, the aim of this research is to present a critical review concerning the Facility Condition Index (FCI): one of the most acknowledged financial indicators (Lavy, Garcia, & Dixit, 2010, 2014a, 2014b. The FCI allows to quantify in a scale from 0 to 100 (where 0 represents the best value) the condition of an asset, based on the expense dedicated to maintenance operations, calculated as the ratio between the cost of Deferred Maintenance (DM) over the Current Replacement Value (CRV). The indicator can be applied at different scales: from a single component to an entire building in a portfolio . The research scope concerns the extensive analysis of international literature, carried out through an investigation on most known scientific literature archives (e.g. Scopus, Web of Science, Science Direct, Google Scholar). Through this analysis, 42 articles have been selected. These documents have been exploited to identify the most acknowledged definitions of the FCI, and to investigate possible drawbacks and advantages arising from its use. The underpinning question concerns the validation of an assets' evaluation procedure based on the calculation of few KPIs, able to represent crucial issues related to the management of the built environment . The need to develop and refine this evaluation procedure comes from some considerations on the contemporary real estate market, more and more characterised by the need for efficiency and interoperability among stakeholders. Monitor of performances is a key issue to achieve better levels of asset management (Pärn et al., 2017;Cable & Davis, 2005), within the context described above. Although, this is often hindered by fragmentation and peculiarity of information concerning the building process, characterised by a great number of actors (Eastman, Teicholz, Rafael, & Kathleen, 2011). PAS 1192-2: 2013(BSI, 2013 gives a common framework to be adopted in order to better coordinate and systematise organisations' information flow in the asset management procedures. Moreover, ISO 55000 series (ISO, 2014), defines performance measurement as a central factor to be considered for management of organisation' tangible goods, balancing opportunities, risks and expenses.
Given the above, starting from the standard definition of the FCI calculation methodology, issues related to the optimisation of expense for maintenance in a condition of scarce resources are tackled in the last section of this article. Figure 1 describes main methodological steps accomplished. In this research the international literature concerning the FCI have been studied. The FCI is one of the most used indicators, describing the performance related to expense for maintenance operations on assets. A recognition of the literature in performance measurement based on economic metrics has been done first.

Research methodology
Main methodological steps (in Figure 1) are: -the bibliographic research, concerning the research and selection of references in most recognised databases; -the bibliometric analysis, concerning network analysis, trend analysis and categorisation by topic, typology of reference and type of FCI formula; -the critical review on the FCI.

Network analysis
In order to provide a wider picture of literature on KPIs for asset and facility management, a network analysis has been carried out (Figure 2 and Figure 3).
For this purpose, references from Scopus and Web of Science database have been retrieved, and the VOSviewer software has been used for full counting of co-occurrence of keywords in references (Perianes-Rodriguez, Waltman, Jan, & Eck, 2016). This software allows to identify the intensity of cited keywords (dimension of the circle) in a selected sample, in terms of number of co-occurrence and strength of connections among keywords (line width). For space reasons, here are presented the network analyses carried out on Scopus database. An extensive extraction of articles concerning KPIs in asset and facility management has been done first ( Figure 2).
Afterwards, a more detailed analysis among publications specifically on FCI has been accomplished ( Figure 3). Among the 134 references analysed, "key performance indicators", "benchmarking"; "maintenance", "construction industry" are the most recurrent terms. This suggests that a conspicuous literature have been published on these topics. Nevertheless, it is appropriate to underline that Figure  2 shows that "maintenance" is connected to the "budget control" keyword only through a weak connection.
Moreover, "budget control" keyword intensity is rather low, suggesting a scarce literature production addressing the issue of budget control in asset and facility management. For the representation in Figure 2, a minimum number of occurrence equal to 5 has been set. Among the 121 keywords matching the parameter, FCI does not appear. Therefore, further bibliographic researches have been carried out, looking for reference in Scopus directly connected to FCI. The research returned only 18 articles on which a network analysis of keywords has been accomplished, setting the minimum occurrence threshold equal to 2. Figure 3 represents the results of this further analysis. The results confirm that FCI is related to strategic asset and facility management issues, despite the sample of references is very scarce. Therefore, the literature review has been extended to further databases not only connected to academia. While the articles have been retrieved mainly from Scopus, Web of Science, Science Direct and Google Scholar, they have been registered by year, author, provenance, typology of reference, type of dissertation and type of FCI described. Following this approach, it was possible to categorise the references as shown in Figure

Extended literature review
Google Scholar have been used to deepen the literature review on the Facility Condition Index. For this purpose, an additional set of keywords have been selected. The most frequent are: "Facility Condition Index", "Facility assessment", "deferred maintenance", "current replacement value", "KPI for asset management", "economic performance measurement", "economic real estate metrics". This is not an exhaustive list, but it is sufficient to outline the field of study. In details, 42 main references have been selected. After the trend analysis in Figure 4, demonstrating a slight increase in publications since the first time FCI appeared in literature, in Figure 5 references are classified according to the main topic discussed categorized according to the following four criteria: -"KPI": is a categorisation for references dealing with FCI only in terms of description of the metric and its use; -"CA strategy" refers to references dealing also with a Condition Assessment (CA) methodology tied to the computation of the FCI; -"AM strategy" concerns those which address the issues related to the definition of a wider Asset Management (AM) strategy; -"BIM process" concerns papers considering FCI as a metric to be integrated in a Building Information Modelling (BIM) process. References have also been categorised by country. Despite Figure 5 and Figure 6 represent the correlation with two different criteria, in both the charts the most frequent provenance is from USA. This happens because Th e classifi cation in Figure 6, allows to verify that many references are not completely connected to academia, but rather to public institutions, which exploit the metric to implement the asset management strategy. Th is outcome is confi rmed by the low sample availability for network analysis as well.
Despite some grey literature has been used (institutional reports), white papers, corporate reports and, in general, technical documentation have been neglected for the classifi cation, since they cannot be considered as impartial sources to be exploited for the literature review.

FCI and derived metrics
Th e following step concerns a more detailed analysis of the contents of the references, namely an investigation on the defi nition of FCI formula. Despite most of the contributions assume as starting point for discussion the basic version given by National Association of College and University Business Offi cers (NACUBO) and Rush (1991), it is possible to spot some cases in which this version is modifi ed to address specifi c issues. Th e somehow hidden modifi cation may cause some biases in calculation of the fi nal score and the clear loss of standardisation. Contributions have been classifi ed according to the version of FCI considered. Figure 7 presents the results of this analysis. Even in this classifi cation, references have been organised by country, showing a similar result as the previous bibliometric analysis. It is worthwhile, at this stage, to specify through Table 1 the criteria used for classifi cation and related references. Table 1 highlights that in most of the references analysed, both in research and in institutional reports, an ad- justed version of the indicator is proposed, because of the high level of simplification of the basic form. Even though simplicity is probably one of the factors that makes the indicator so effective, it can be seen as a limitation to be balanced through corrective parameters or through the combination with other indicators. These corrective parameters widen the range of phenomena gathered by the indicator.

FCI calculation methods
The indicator, published in its first version by NACUBO and Rush (1991) thanks to the collaboration with Applied Management Engineering (AME) (Kaiser, 1993), was initially defined with the expression (1): where: DM is the cost of Deferred Maintenance; CRV is the Current Replacement Value.
The DM is defined as the monetary expenditure for the deferred maintenance operations and renewals on components, systems or the whole building, that should have been accomplished in a given period, but have been delayed. The Department of the Interior (DoI) defines DM as comprehensive of type of maintenance described in Table 2. On the other hand, CRV is defined as the replacement value of components, systems or the whole building (U.S. Department of the Interior, 2008b).
The result of the calculation ranges from 0 to 100 (where 0 is the best condition) and is usually organised in three or four levels, whose thresholds can slightly vary according to the type of asset under analysis: -good: 0−5%; -fair: 5−10%; -poor: 10−30%; -critical: 30−100%. Worth to be notice that DoI, describes the calculation methodology for FCI in its institutional documents, though for the actual computation, it refers to the use of   (Rush, 1991;Kaiser, 1993;Fagan & Kirkwood, 1997 Since the definition of its first version in Eq. (1), the indicator had been revised and other versions have been proposed. Recent versions of the FCI mainly differ in the calculation of the DM including, for instance, interventions accomplished to correct obsolescence due to the lack of compliance with codes. Other DM calculation methodologies are specific for the evaluation of historical assets or infrastructures.
The calculation of the performances related to the expense for maintenance operations can be adjusted in order to describe the overall performance of the asset or portfolio. Thus, the FCI can be tied to technical, functional and environmental parameters. The DoI, for instance, adopts a combination of indicators derived from the FCI to measure performances of high valuable and special assets. The DoI portfolio is composed by conventional buildings, as well as by infrastructures and heritage buildings. Therefore, the guideline for the implementation of the Site-Specific Asset Business Plan (U.S. Department of the Interior, 2005bInterior, , 2008b, encourages the use of a combination of three variations of the FCI (2)(3)(4): -the FCI dm , as defined in Rush (1991) The equation is used for the calculation of the Condition Index CI comp (6): The CI comp, thus, can provide a wider evaluation of economic impact of maintenance for high valuable assets. The extended version of the FCI described above is just one of the many that can be found in literature. Nevertheless, it can be considered as a good example of how this metric has been modified over the time.
Other cases in which FCI is combined with other metrics can be identified in literature. To relate the economic performance to the organisation's asset policy, the DoI encourage the utilisation of the CI comp or the simple CI, coupled with the Asset Priority Index (API) (U.S. Department of the Interior, 2005a): an indicator describing the relevance of the asset according to the goals of the owning organisation (U.S. Department of the Interior, 2005b. The API is a metric exploited also by private industries and other government agencies such as the National Aeronautics and Space Administration (NASA), the US Navy and the US Air Force (Mills, 2001;NASA, 2003). This approach underlies a more comprehensive calculation of the FCI, which allows to comprehend not only the performance in terms of expense to contrast the degradation phenomena, but also obsolescence process. Moreover, also issues related to the strategic value of the asset are considered.
Some other indicators, derived from the FCI, can be found in literature. Lavy et al. (2014), for instance, after discussing most effective KPIs for facilities management, identify an array of few relevant indicators for measurement of physical, functional, user satisfaction and financial performances. For the latter type of measurement, they propose the Maintenance Efficiency Indicator (MEI) (7) and the Replacement Efficiency Indicator (REI) (9). The MEI is used for assessment of efficiency of maintenance expenditure in a given year. It is calculated considering that it is not always possible to have access to all the resources needed for covering the total amount of deferred maintenance, thus a target FCI (calculated as the CI, namely 1-FCI) must be defined. A MEI of 100% means that the precise amount needed for achieving the target CI is spent. A MEI of 80% correspond to an underspending of resources. Conversely, a MEI equal to 110% indicates an overspending. Replacement is not considered in the calculation of this metric. where: The REI, is a metric giving a dimension to the expenditure related to substitutions of components. REI below 100% means that amount spent for replacement is not enough. On the other hand, a REI greater than 100% indicates that it has been spent more than necessary. It must be considered that components must not be replaced each year, thus REI is not a continuous function, if it is considered over the facility's lifespan. This indicator describes the trend for substitution of systems at the asset level, since it considers the overall expenditure for substitution of all the components of a building. An analogous calculation at the single component level would not be significant. where: . Exp Cost ∑ is the actual expense for replacement of components and systems; . Exp Cost ∑ is the total replacement cost of expired components and systems in the same year.
MEI and REI measure how efficiently maintenance budget is spent over the years. Some other examples of combined indicators derived from FCI can be found in literature, as highlighted in in Table 1. Nevertheless, MEI and REI can be considered powerful metrics to evaluate the effectiveness of expenses for maintenance interventions (Shohet, Kot, & Karako, 2008;Lavy et al., 2014).

Computing deferred maintenance and current replacement value
Some insights related to FCI and its utilisation can be highlighted. At first, it can be stated that this indicator should be employed jointly with a well-defined CA procedure, specifying the set of operations to be accomplished (Lavy, Garcia, Scinto, & Dixit, 2014;U.S. Department of the Interior, 2008b, 2008a. Since FCI is a KPI that can be exploited for the evaluation of either single components or the entire building, the assessment phase is crucial to avoid biases in the decision-making process. Moreover, the calculation methodology can vary according to the objective to be achieved (Federal Facilities Council, 2001), though a standardised set of operations to compute the DM and the CRV are suggested by the US Department of the Interior. This institution adopts the FCI as a primary indicator to carry out the asset management policy (U.S. Department of the Interior, 2008b, 2008a) on its huge portfolio including real properties for a total value of more than 240 billion (U.S. Department of the Interior, 2017a). The Federal Facilities Council (2001) suggests two main assessment methodologies: the Condition Assessment Survey and the Total Life Cycle Cost Method, despite some others can be found in literature (Cable & Davis, 2005;Bello & Loftness, 2010). The former method encompasses a series of periodic inspections carried out to determine the current condition of properties, plants and equipment. The latter gives as outcome a forecast used to compare actual expense in maintenance to the yearly budget. Altogether, a clear explanation about algorithms to be used in calculation cannot be found (Roberts, 2009) since the methodologies are mainly explained in terms of their scope and final goals. The most critical issue seems to be the aggregation of FCI values from the components to the whole asset (Selman, 2003). Moreover, methodology for economic definition of Deferred Maintenance (DM) and Current Replacement Value (CRV) can vary remarkably because of differences in assessment methodologies adopted, composition of costs in local pricelists, regional factors and discrepancies in national labour markets (U.S. Department of the Interior, 2008b).

Possible drawbacks in FCI application
In its basic form, the FCI measures the performance of a building or a component in terms of the ratio be-tween DM over CRV. This ratio does not indicate the magnitude of deferred maintenance interventions to be executed, since the FCI is mainly led by the value of the denominator. For instance, two analogous low cost (deferred) maintenance interventions on two components characterised by very different CRV values determine substantial differences in the FCI calculation. To be more specific, the FCI will be more critical for the component featured by the lower CRV value. Nevertheless, that simple intervention could be highly critical for the component featured by the highest replacement cost. Therefore, a simple FCI calculation could not be representative of the criticality of the maintenance operation to be carried out.
where: is the ratio between CRV values. Figure 8 and Eq. (10) show how the ratio of the FCI of two components changes according to variations in the ratio between the deferred maintenance costs (DM ratio) and the current replacement values (CRV ratio) of the components. It can be seen that for a given DM ratio the FCI ratio increases with a linear trend as well as the CRV ratio. For a given CRV ratio the FCI ratio decreases following a hyperbolic trend with the DM ratio increasing.
Accordingly, it can be stated that the FCI is a metric lead by the replacement cost of the components: a critical issue to be considered in decision-making, especially in a scarce resource condition. This characteristic may cause unreliability of comparisons among assets, a particularly relevant issue when a maintenance prioritisation strategy is being defined. However, FCI is almost always related to an asset management strategy, driving strategic investment decisions (Bello & Loftness, 2010;Rashedi & Hegazy, 2016;. Moreover, the FCI in its simple version does not consider other possible issue related to non-economic value, though many other aspects should be considered (technical performances, functions, relevance for community, user and owners, etc.). Thus, it is worthwhile to adjust the FCI through some parameters which allow to catch more extensive issues related to asset management. In literature some studies that relate FCI to other metrics can be found Roberts, 2009;Bello & Loftness, 2010;Lavy, 2013;Parsons, 2013;Determan, Akers, Williams, Hohmann, & Martin-Dunlop, 2015;Marzouk & Awad, 2016), though a quantitative methodology to directly compute the dependency among performances and economic value of facilities is not clear.
Thanks to the literature review, it is possible to state that the FCI is mainly applied at the asset level. For instance, it has been widely exploited by US public department in management of wide real estate portfolios and by US real estate organisations (AAPPA, 2000;IFMA, 2008;Magellan Consulting, 2011;NASA, 2003;U.S. Department of the Interior, 2008b). When dealing with components instead of the whole buildings, the level of detail must be carefully defined. It must be considered, for instance, that when a lower breakdown level is reached, a further complexity concerning bi-stable components (e.g. bulbs, switchers, etc.) will be faced. In this case the FCI value can assume only two values, namely 0 or 1, not giving the magnitude of the degradation trend for those components. Moreover, Uzarski and Grussing (2008) suggested to use the FCI together with a strong CA procedure and a precise calculation methodology based on updated price lists and standardised maintenance costs archives.
The FCI can also be effective in the evaluation of facility management companies (Yik & Lai, 2005) if in the negotiation phase, a Service Level Agreement (SLA) based on the indicator is defined. In this case, the trend of FCI over the years should be registered. It should be noticed that in a FM contract the agreed FCI can be higher than zero because of a strategic decision. Conversely, in a scarce resources circumstance, the target FCI can be used for optimisation of maintenance expenses.

Conclusions
To conclude, the critical review presented in this paper identifies the main characteristics of the FCI, through a study based on a methodological approach organised in three subsequent phases: -bibliographic research, -bibliometric analysis, -critical review. However, it can be considered as a first step in the wider research on asset management processes. A conspicuous literature regarding the performance measurement of buildings has been investigated through the literature review, despite only few references are explicitly referred to FCI. In spite of that, large institutions as the US Government, exploit this metric for the definition of the asset management policy.
Concerning the calculation of the metric, it must be considered that formulas described in this article are only some of those that can be found in literature. Nevertheless, they are representative, since they are amongst the most employed by academics and professionals. A widely acknowledged computation methodology for FCI has not been defined yet, since from the first version, the metric has been revised and adapted. A standardised calculation methodology along with a clear condition assessment procedure could allow to assess portfolios more effectively and guide expenditure in maintenance, even in a scarce resource condition.
Eventually, considering the AM business process, KPIs could be categorised in two domains: -the core domain, comprehending the set of policies and strategies carried out to achieve the main goal of the company, linked to its primary objectives; -the non-core domain, defined as the set of policies and strategies to be accomplished in order to facilitate and foster the companies' core business. It is clear that the FCI allows to measure the performance of non-core issues. In literature, only few attempts to link non-core performances to core business activities can be identified. An example can be found in the research regarding the improvement of teaching activities in well maintained school buildings. These studies show how a better teaching environment can affect cognitive capabilities of scholars (Determan et al., 2015;Bello & Loftness, 2010;Roberts, 2009). These approaches stem from a qualitative performances evaluation and give a fair result in terms of correlation between non-core performances measurement and achievement of core objectives. Therefore, it could be worthwhile to define a quantitative methodology to catch the relationship between core and non-core performances.
To summarize, it can be stated that the FCI is a reliable KPI, despite sources in literature agree on some limitations: -it is highly influenced by the CRV, therefore it must be related to other metrics, in order to catch, for instance, the criticality of the components; -it is not possible to spot a widely acknowledged or standardised calculation methodology; -it cannot measure by itself the core performances of a company. These issues will be tackled in further studies.